Model comparison
Qwen2.5-Coder-32B vs Qwen3.5 27B
Qwen3.5 27B is the stronger model overall, scoring 41.4 to 32.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
Alibaba (Qwen)Summary
- They share 12 benchmarks with published results for both. Qwen2.5-Coder-32B scores higher in 0 categories and Qwen3.5 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5 27B leads 59.5 to 41.8.
- Qwen2.5-Coder-32B is cheaper at $0.66 / $1 per million input/output tokens, against $0.30 / $2.40 for Qwen3.5 27B.
- Qwen3.5 27B accepts more context: 262K tokens versus 33K.
Side by side
| Qwen2.5-Coder-32B | Qwen3.5 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 32.9 | 41.4 |
| Released | 2024-09-18 | 2026-02-23 |
| Weights | Open | Open |
| Context window | 33K | 262K |
| Max output | 29K | 66K |
| Input $ / M tokens | $0.66 | $0.30 |
| Output $ / M tokens | $1 | $2.40 |
| Results tracked | 31 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Qwen2.5-Coder-32B: 22.4 (#328), Qwen3.5 27B: 38.6 (#159)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1276 | 1427 |
| SWE-bench Verified (bash only) | 9% | — |
| Aider Polyglot | 16.4% | — |
| LMArena WebDev | — | 1357 |
| WeirdML | — | 39.5% |
| BigCodeBench Instruct | 49% | — |
| LiveBench Coding | 56.9% | — |
| BigCodeBench Complete | 58% | — |
| ALE-Bench | — | 349.45 |
| HumanEval+ | 87.2% | — |
| MBPP+ | 77% | — |
Agentic & Tool Use Not comparable
Qwen2.5-Coder-32B: —, Qwen3.5 27B: —
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| Vending-Bench 2 | — | 201.98 |
Reasoning Qwen3.5 27B leads
Qwen2.5-Coder-32B: 20.4 (#220), Qwen3.5 27B: 26.6 (#116)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1251 | 1414 |
| NYT Connections (extended) | — | 47.9% |
| Thematic Generalization | — | 45.5% |
| LiveBench Reasoning | 42.1% | — |
| DTBench | — | 82.4% |
| LiveBench Data Analysis | 49.9% | — |
| LMCA | — | 34% |
| Epoch Capabilities Index | 119.41 | — |
| HellaSwag | 83% | — |
| LiveBench | 46.2% | — |
| WinoGrande | 80.8% | — |
Math Qwen3.5 27B leads
Qwen2.5-Coder-32B: 30.8 (#203), Qwen3.5 27B: 36.8 (#103)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1251 | 1428 |
| MathArena Final-Answer Competitions | — | 56.7% |
| LiveBench Math | 46.6% | — |
| GSM8K | 93% | — |
Knowledge Qwen3.5 27B leads
Qwen2.5-Coder-32B: 32.9 (#200), Qwen3.5 27B: 37.4 (#149)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Expert | 1221 | 1428 |
| Vectara Hallucination Rate | — | 12.1% |
| ARC (AI2) Challenge | 70.5% | — |
| MMLU | 79.1% | — |
Multimodal Not comparable
Qwen2.5-Coder-32B: —, Qwen3.5 27B: 38.7 (#56)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1240 |
Multilingual Qwen3.5 27B leads
Qwen2.5-Coder-32B: 38.0 (#230), Qwen3.5 27B: 51.0 (#112)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1205 | 1390 |
| LMArena Chinese | 1222 | 1478 |
| LMArena Russian | 1228 | 1389 |
| LMArena French | — | 1409 |
| LMArena German | — | 1393 |
| LMArena Japanese | — | 1345 |
| LMArena Korean | — | 1359 |
| LMArena Spanish | — | 1408 |
Instruction Following Qwen3.5 27B leads
Qwen2.5-Coder-32B: 61.6 (#240), Qwen3.5 27B: 73.7 (#114)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1223 | 1393 |
| LiveBench Instruction Following | 58.7% | — |
Long Context Qwen3.5 27B leads
Qwen2.5-Coder-32B: 37.9 (#204), Qwen3.5 27B: 43.1 (#103)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1251 | 1413 |
Writing & Preference Qwen3.5 27B leads
Qwen2.5-Coder-32B: 41.8 (#235), Qwen3.5 27B: 59.5 (#106)
| Benchmark | Qwen2.5-Coder-32B | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1230 | 1409 |
| LMArena Creative Writing | 1174 | 1362 |
| LMArena Multi-Turn | 1222 | 1410 |
| LiveBench Language | 23.3% | — |
Frequently asked questions
Is Qwen2.5-Coder-32B better than Qwen3.5 27B?
Qwen3.5 27B is the stronger model overall, scoring 41.4 to 32.9 on the Noometry Index.
Which is cheaper, Qwen2.5-Coder-32B or Qwen3.5 27B?
Qwen2.5-Coder-32B is cheaper. It lists at $0.66 per million input tokens and $1 per million output tokens; Qwen3.5 27B lists at $0.30 and $2.40.
Is Qwen2.5-Coder-32B or Qwen3.5 27B better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.6 versus 22.4 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 27B does, with 262K tokens against 33K.
How many benchmarks do Qwen2.5-Coder-32B and Qwen3.5 27B share?
12 benchmarks have published results for both models. Qwen2.5-Coder-32B has 31 scored results on Noometry and Qwen3.5 27B has 28.